<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: Do I need to standardize the variables while using proc glmselect LASSO or adaptive LASSO? in New SAS User</title>
    <link>https://communities.sas.com/t5/New-SAS-User/Do-I-need-to-standardize-the-variables-while-using-proc/m-p/536104#M6581</link>
    <description>&lt;P&gt;&lt;A href="https://go.documentation.sas.com/?docsetId=statug&amp;amp;docsetTarget=statug_glmselect_details10.htm&amp;amp;docsetVersion=15.1&amp;amp;locale=en" target="_self"&gt;Here is the documentation.&amp;nbsp;&lt;/A&gt;&lt;/P&gt;
&lt;P&gt;I interpret the doc to say that covariates are internally centered and scaled during the LASSO process. However, the final parameter estimates are provided for the original variables.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;If you want to run the experiment, you could use PROC STDIZE to standardize your regressors and run the program twice, once on the original data and once on the standardized data. The selected effects should be the same. I did a similar experiment in the article &lt;A href="https://blogs.sas.com/content/iml/2018/08/22/standardized-regression-coefficients.html" target="_self"&gt;"Standardized regression coefficients."&lt;/A&gt;&lt;/P&gt;</description>
    <pubDate>Sat, 16 Feb 2019 01:18:11 GMT</pubDate>
    <dc:creator>Rick_SAS</dc:creator>
    <dc:date>2019-02-16T01:18:11Z</dc:date>
    <item>
      <title>Do I need to standardize the variables while using proc glmselect LASSO or adaptive LASSO?</title>
      <link>https://communities.sas.com/t5/New-SAS-User/Do-I-need-to-standardize-the-variables-while-using-proc/m-p/535894#M6548</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Does anyone know whether "proc glmselect" will automatically standardize all the variables while running LASSO and adaptive LASSO? "Standardize" means demean the variable and scale it by the standard deviation.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thank you!&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Best,&lt;/P&gt;&lt;P&gt;Yutong&lt;/P&gt;</description>
      <pubDate>Fri, 15 Feb 2019 16:24:21 GMT</pubDate>
      <guid>https://communities.sas.com/t5/New-SAS-User/Do-I-need-to-standardize-the-variables-while-using-proc/m-p/535894#M6548</guid>
      <dc:creator>YutongHU1</dc:creator>
      <dc:date>2019-02-15T16:24:21Z</dc:date>
    </item>
    <item>
      <title>Re: Do I need to standardize the variables while using proc glmselect LASSO or adaptive LASSO?</title>
      <link>https://communities.sas.com/t5/New-SAS-User/Do-I-need-to-standardize-the-variables-while-using-proc/m-p/536104#M6581</link>
      <description>&lt;P&gt;&lt;A href="https://go.documentation.sas.com/?docsetId=statug&amp;amp;docsetTarget=statug_glmselect_details10.htm&amp;amp;docsetVersion=15.1&amp;amp;locale=en" target="_self"&gt;Here is the documentation.&amp;nbsp;&lt;/A&gt;&lt;/P&gt;
&lt;P&gt;I interpret the doc to say that covariates are internally centered and scaled during the LASSO process. However, the final parameter estimates are provided for the original variables.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;If you want to run the experiment, you could use PROC STDIZE to standardize your regressors and run the program twice, once on the original data and once on the standardized data. The selected effects should be the same. I did a similar experiment in the article &lt;A href="https://blogs.sas.com/content/iml/2018/08/22/standardized-regression-coefficients.html" target="_self"&gt;"Standardized regression coefficients."&lt;/A&gt;&lt;/P&gt;</description>
      <pubDate>Sat, 16 Feb 2019 01:18:11 GMT</pubDate>
      <guid>https://communities.sas.com/t5/New-SAS-User/Do-I-need-to-standardize-the-variables-while-using-proc/m-p/536104#M6581</guid>
      <dc:creator>Rick_SAS</dc:creator>
      <dc:date>2019-02-16T01:18:11Z</dc:date>
    </item>
  </channel>
</rss>

